Chapter 4 Simulation Results
4.3 Low-Complexity Time Domain Channel Estimation Results
In this section, we evaluate the performance of the low-complexity LS channel estimator (in the time domain). Here, we assume that the exact positions of channel taps are perfectly known. Figure 4-7 shows the results. We can see that the low-complexity LS channel estimator has very good performance; it can closely approach the ideal case. In the same figure, we can also see that the LAQ and slide-window LAQ methods exhibit an error floor phenomenon in high SNR. In order to reduce the computational complexity, we can take fewer pilot signals in the LS calculation. According to the IEEE802.16e specification, when the symbol size is 2048, there are 240 pilot tones. Figure 4-8~Figure 4-9 shows the BER performance for the cases with different number of used pilot tones. From Figure 4-8, we can see that since the channel delay spread is not particularly large, the performance is good even when the number of pilot tones used is small. From Figure 4-9, we can see the performance becomes affected when the delay spread is large and fewer pilot tones are used.
Figure 4-7 BER performance comparison for various estimation methods (delay spread > 85TS)
5 10 15 20 25 30 35 40 10-4
10-3 10-2 10-1 100
Np=60 Np=120 Np=240 perfect
Figure 4-8 BER performance vs. number of used pilot signals (55TS< delay spread < 85TS)
5 10 15 20 25 30 35 40
10-4 10-3 10-2 10-1 100
Np=60 Np=120 Np=240 perfect
Figure 4-9 BER performance vs. number of used pilot signals (delay spread > 85TS)
4.4 Joint Time and Freq Domain Channel Estimation Results
As shown in Section 4.3, if the channel tap positions are known, the performance of the proposed time domain channel estimator can closely approach to that of the ideal case even when the channel delay spread is large. The key becomes how to locate the channel-tap positions. Figure 4-10 shows an example of the preliminary channel estimation result. In this example the positions of the channel are [1 2 3 8 10 11]. With the proposed channel tap identification method, the located taps are shown in Figure 4-11. Figure 4-12 shows the BER performance for the proposed joint time and frequency domain channel estimator. It is apparent that the proposed method works very well.
0 5 10 15 20 25 30 35 40 45
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4
Ts
magnitude
Figure 4-10 Preliminary channel estimation result with pilots
0 5 10 15 20 25
4.5 Time-Variant Channel Estimation Results
In Section 4.2, 4.3, and 4.4, we assume that the channel is quasi-stationary, which means that the channel is time-invariant in an OFDM symbol. In this section, we will consider the time-variant channel scenario. We will first conduct the channel estimation with the proposed method and then compensate for the ICI effect. Figure 4-13 and 4-14 shows the BER performance (for different mobile speed) without ICI cancellation. As we can see, the higher the mobile speed, the large the performance degradation it will result. We use the proposed channel estimator to estimate , and then construct the time-variant channel. Figure 4-15, 4-16, ands 4-17 show the estimation of a certain channel tap for different mobile speeds. When the mobile speed is 50 km/hr, the estimated time-variant tap is overlapped to the actual time-variant channel tap. However, as the mobile speed becomes higher, the estimation result becomes degraded. With the estimated channel, we can conduct ICI cancellation. Figure 4-18 shows the BER performance for ICI cancellation with the MMSE method. We can see that the ICI is eliminated effectively, indicating that the proposed channel estimator does do a good job.
ˆave hk
5 10 15 20 25 30 35 40
2000 4000 6000 8000 10000 12000 14000
0.18
Figure 4-18 BER performance for ICI cancellation by MMSE equalizer
Chapter 5
Conclusions and Future Works
In this thesis, we focus on channel estimation in mobile OFDM systems. We first propose an enhanced LAQ method, improving the channel estimation in 802.16e system. With a sliding-window approach, we can obtain smoother response in the boundary areas of clusters. Simulations show that the performance of the slide-window LAQ method is better than the original LAQ method though its computational complexity is somewhat higher.
We then develop a high-performance yet low-complexity LS channel estimator.
The key idea of the proposed method is to locate and estimate the channel responses at non-zero positions. Due to the number of taps to be identified is significantly reduced, the computational complexity of the LS algorithm is significantly reduced also. With an iterative method, the computational complexity of the channel tap search can be further reduced. Combining both temporal and frequency domain processing, we can obtain a joint time-and-frequency-domain channel estimator.
Simulations show that the performance of the proposed channel estimator can closely approach that of the ideal case in which the channel response is perfectly known.
Finally, we consider the time-variant channel scenario, in which the ICI is introduced. To mitigate the ICI, the channel response has to be known. We extend the proposed channel estimator for time-invariant channels to do the job. Similar to the time-invariant scenario, the proposed method can accurately identify the time-variant
channel even when the mobile speed is high.
In this thesis, we only consider single-input-single-output (SISO) systems. As we known, multiple-input-multiple-output (MIMO) systems can provide much higher throughput. Thus, MIMO OFDM systems are becoming more popular. As a nature extension, we may apply methods proposed in this thesis to the channel estimation problem in MIMO OFDM systems. For ICI cancellation, we only try the MMSE method. However, the MMSE method requires very high computational complexity, is not suitable for implementation. There are various alternatives with much lower complexity. How to combine the channel estimation and ICI cancellation in an efficient way may also serve a potential topic for further research.
Reference
[1]. J. -J. van de Beek, O. Edfors, M. Sandell, S.K. Wilson, and P. O. Borjesson, “On channel Estimation for OFDM Systems,” Proc.Vehicular Technol. Conf., 1995, pp. 815-819.
[2]. Jei Zhu, Wookwon Lee, “A low-complexity channel estimator for OFDM systems in multipath fading channels,” IEEE PIMRC 2004. 15th IEEE international Symposium, vol. 3, pp. 1978-1982, Sept. 2004.
[3]. A. R. S. Bahai and B. R. Saltzberg, Multi-carrier Digital Communications Theory and Applications of OFDM. New York: Kluwer Academic, 1999.
[4]. R. Prasad, OFDM for Wireless Communications Systems. Boston: Artech House, 2004.
[5]. M. Engels, Wireless OFDM System: How to Make Them Work? Boston: Kluwer Academic Publishers, 2002
[6]. IEEE 802.16-2004, “Part 16: Air Interface for Fixed Broadband Wireless Access Systems,” IEEE Std 802.16-2004
[7]. IEEE 802.16e, “Part 16: Air Interface for Fixed and Mobile Broadband Wireless Access Systems,” IEEE Std 802.16e-2005
[8]. M.-H. Hsieh, “Synchronization and channel estimation techniques for OFDM systems,” Ph.D. dissectation, Department of Electronics Engineering, National Chiao Tung University, Hsinchu, Taiwan, R.O.C., May 1998.
[9]. Steven M. KAY, Fundamentals of Statistical Signal Processing : Estimation Theory, Prentice Hall, 1993
[10]. O. Edfors, M. Sandell, J. J. van de Beek, S. K. Wilson, and P. O. BÄorjesson,
“OFDM channel estimation by singular value decomposition,” in IEEE 46th Vehicular Technology Conference, Apr. 1996, pp. 923-927.
[11]. Tao Cui, Chintha Tellambura, “Superimposed Pilot Symbols for Channel Estimation in OFDM Systems,” in Proc. IEEE GLOBECOM’ 05 , vol 4, pp.
2229-2233, Nov. 2005.
[12]. Wan-Yi Lin, Wen-Rong Wu, “Channel Estimation and Multiuser Asychronization Interference Mitigation for IEEE 802.16e System,” National Chiao-Tung University, July 2006
[13]. Y. –H. Yeh, and S. -G. Chen “Reduction of Doppler-induced ICI interference prediction,” IEEE PIMRC 2004. 15th IEEE international Symposium, vol. 1, pp.
653-657, Sept. 2004.
[14]. W. –S. Hou and B. -S. Chen, “ICI cancellation for OFDM Communication Systems in Time-Varing Multipath Fading Channels,” IEEE Transaction on Wireless Communications, vol. 4, no. 5, pp. 2100-2110, Sep. 2005.
[15]. Y. Mostofi and D. C. Cox, “ICI Mitigation for Pilot-Aided OFDM Mobile Systems,” IEEE Transactions on Wireless Communications, vol. 4, no. 2, pp.
765-774, Mar. 2005.
[16]. Ranganathan Krishnan, Tamer Kadous, “Channel Estimation for OFDM Communication Systems,” U. S. Patent 7039001, May 2, 2006.
[17]. 3GPP TR 25.996 ,“Spatial channel model for Multiple Input Multiple Output (MIMO) simulations”, v6.1.0, Sept.2003